Hugo V.

• • Build high-signal AI benchmarks to evaluate coding agents on Spanish-language software-engineering and terminal workflows.

• • Create reproducible Docker-based task environments with localized datasets, bilingual instructions, oracle solutions and deter ministic verifiers.

• • Run Claude-based agents in CLI workflows and analyze execution logs, pass@k results, Unicode/encoding issues and locale specific edge cases for benchmark calibration.

I design, build, and evaluate production-oriented Generative AI systems for enterprise use cases. My work focuses on RAG architectures, document intelligence, AI agents, API-based automation, and cloud-native AI services.

I work mainly with Python, FastAPI, Azure OpenAI, Azure AI Search, Docker, REST APIs, CI/CD pipelines, and observability tools. I have contributed to AI backends that process documents, retrieve relevant information, generate structured outputs, automate internal workflows, and connect LLMs with business systems through APIs.

I also work on evaluation and reliability aspects of AI systems, including prompt design, retrieval quality, structured output validation, debugging, monitoring, and failure analysis. My focus is to build AI solutions that are not only functional prototypes, but reliable, maintainable, measurable, and ready for real production environments.

This experience also strengthened my ability to communicate technical decisions clearly, work with enterprise constraints, and deliver AI solutions that create concrete business value.

• Full development of a mobile application integrating artificial intelligence to automate real-time voice interactions.

• Use of SwiftUI, Kotlin, and FastAPI to ensure a smooth and scalable experience.

• Design, fine-tuning, and deployment of AI models tailored to user needs.

• End-to-end project management: architecture, cloud integration, strategic partnerships, and generating investor interest.